This PDF is a selec on from a published volume from the Na ... · This PDF is a selec on from a...

This PDF is a selec on from a published volume from the Na onal Bureau of Economic Research

Volume Title: Fiscal Policy a er the Financial Crisis

Volume Author/Editor: Alberto Alesina and Francesco Giavazzi, editors

Volume Publisher: University of Chicago Press

Volume ISBN: 0‐226‐01844‐X, 978‐0‐226‐01844‐7 (cloth)

Volume URL: h p://www.nber.org/books/ales11‐1

Conference Date: December 12‐13, 2011

Publica on Date: June 2013

Chapter Title: Fiscal Rules: Theore cal Issues and Historical Experiences

Chapter Author(s): Charles Wyplosz

Chapter URL: h p://www.nber.org/chapters/c12656

Chapter pages in book: (p. 495 ‐ 525)

495

12Fiscal RulesTheoretical Issues and Historical Experiences

Charles Wyplosz

12.1 Introduction

The European sovereign debt crisis is an unwelcome reminder that no country can ignore the requirement of fi scal discipline. It should also clarify many issues on the nature of fi scal discipline and on the ways to achieve it. We knew it all, but a few aspects have been made more concrete.

The crisis illustrates how slowly fi scal discipline can assert itself. Govern-ments can run budget defi cits for years, even decades, before fi rst facing the wrath of fi nancial markets and next facing their emergency lenders. This is illustrated for the Organization for Economic Cooperation and Develop-ment (OECD) countries by table 12.1, which shows the percent of years when a country has run a budget defi cit since 1960. This should happen about half of the time in a disciplined- government country, and indeed this is what is found for Denmark, New Zealand, and Sweden (Norway being a clear outlier because of its intergenerational saving into the Petroleum Fund). For all other countries except Finland, defi cits have occurred in four years out of fi ve (or more), with two countries (Italy and Portugal) achieving a perfect 100 percent. The table also shows that the last time when Austria, Greece, and France achieved a surplus was before the fi rst oil shock. Thus

Charles Wyplosz is professor of international economics at the Graduate Institute of Inter-national and Development Studies in Geneva, where he is director of the International Centre of Money and Banking Studies (ICMB).

Without implicating them, I am grateful to Alberto Alesina, Frits Bos, Xavier Debrun, Fran-cesco Giavazzi, and Lucio Pench for their comments on earlier versions. I also benefi ted from comments from participants at the NBER conference in Cambridge in July 2011 and at the NBER- Bocconi conference in Milan in December 2011. I acknowledge with gratitude fi nancial support from the European Commission under the PEGGED program. For acknowledgments, sources of research support, and disclosure of the author’s material fi nancial relationships, if any, please see http: // www.nber.org / chapters / c12656.ack.

496 Charles Wyplosz

defi cits can be the rule, with few exceptions, for fi fty years or more while these countries record AAA or close ratings.1

The euro area countries currently under International Monetary Fund–European Union (IMF–EU) programs (Greece, Ireland, and Portugal) have all expressed shock at fi nding themselves under market pressure. This may refl ect a misleading conviction that debt crises only occur in developing or emerging market countries. It also refl ects that years and decades of lenient appraisal by the fi nancial markets and rating agencies can come to a sur-prisingly abrupt end. Sudden stops have long been seen as a very serious threat, quite possibly refl ecting self- fulfi lling phenomena. Debts can grow unnoticed until they get noticed. They represent the kind of vulnerability that gives rise to self- fulfi lling crises.

Three unmistakable implications follow. First, fi scal discipline is not a year- by- year concept, in sharp contrast with the prescription of the Stability and Growth Pact. It is a medium- to long- term characteristic, which may allow for signifi cant temporary slippages along with eventual offsetting sur-pluses. Second, a good track record is not sufficient to rule bad equilibria.2

Table 12.1 Percent years of defi cit over 1960–2011

Australia Austria Belgium Canada Germany

Percent (%) 80 82 96 76 78Last surplus 2008 1974 2006 2007 2008

Denmark Spain Finland France UK

Percent (%) 48 78 20 90 84Last surplus 2008 2007 2008 1974 2001

Greece Ireland Italy Japan Netherlands

Percent (%) 80 80 100 68 88Last surplus 1972 2007 1992 2008

Norway New Zealand Portugal Sweden US

Percent (%) 4 46 100 42 92Last surplus 2011 2008 2008 2000

Sources: Economic Outlook, OECD; and Eichengreen and Wyplosz (1998) for older data.Note: Sample starts later for Australia (1962), Canada (1961), Spain (1962), and Portugal (1977).

1. It can be argued that primary budget balances offer a more accurate description of gov-ernment behavior. Time series are shorter (going back to 1970 at best). They provide a similar picture, although highly indebted governments do much better, and are available from the author upon request. It remains that the budget laws, voted by parliaments, highlight the overall balance, which represents what policymakers explicitly decide upon.

2. Both Spain and Ireland achieved large debt reductions in the years leading to the fi nancial crisis. Yet, they had to deal with the consequences of their housing price bubbles—they were not able to reassure the fi nancial markets that they had the ability to eventually close their defi cits.

Fiscal Rules: Theoretical Issues and Historical Experiences 497

A solid budgetary framework is a necessary, but not sufficient, condition because private debts can become public debts in crisis situations. Finally, the policy dominance issue is a concern that no central bank can escape, no matter how independent it may be.

Von Hagen and Harden (1994) take a wide view of what constitutes a budget process, including arrangements within the government. The thrust of their analysis is that, year in and year out, fi scal discipline is achieved either when the Finance Minister is given enough authority to control the process, or when the political forces that support the government agree to adequate contracts. Hallerberg, Strauch, and von Hagen (2009) provides some supporting empirical evidence in the case of European countries. This chapter takes a different, but related approach. It directly looks at two types of processes: fi scal numerical rules and fi scal institutions. This is in line with the recent literature (see, e.g., Kopits and Symanski 2001; Wyplosz 2005; and Debrun, Hauner, and Kumar 2009).

Fiscal rules come in a large variety of forms but they share the character-istic of imposing numeral norms. These norms can concern the budget bal-ance, public spending, or government revenues. The limitations of rules are well known (Kydland and Prescott 1977): because they are fundamentally arbitrary and noncontingent, rules must sometimes be suboptimal, which creates a serious time- consistency problem. This argues in favor of institu-tions, meaning formal arrangements that are designed to prescribe actions optimally designed to respond to unforeseen contingencies. Unfortunately, the conditions required for fi scal institutions to be effective are rarely met in practice.

The next section examines the theoretical foundations for fi scal rules and their empirical relevance. Section 12.3 presents the theory behind the need to adopt restraints on the budgetary process. Section 12.4 then describes the various forms of rules. Section 12.5 considers a number of arrangements and draws policy implications. The last section concludes.

In this chapter I defi ne restrictively fi scal rules as numerical rules (Kopits and Symanski 2001). This defi nition excludes institutions; that is, formal procedures that do not rely on quantitative restraints but that shape the budgetary process. Both rules and institutions have a role to play because it can be valuable to constrain policymakers. Indeed, policymakers display a defi cit bias for fundamental reasons, which are examined in section 12.3.

12.2 The Defi cit Bias: Theory and Evidence

In the absence of any defi cit bias, we would observe budgets to be alter-natively in defi cit and in surplus depending on economic and / or political conditions. These fl uctuations would be mainly driven by business cycles when fi scal policy is run countercyclically, as should be. The frequency of balance fl uctuation could also be longer when governments borrow to invest

498 Charles Wyplosz

during a catch- up phase, or because of war efforts, or else in the aftermath of a fi nancial crisis that has led the government to bail some of its banks out. Fiscal discipline is present when, over the long run, the debt- to- GDP ratio is stationary.3 The point is illustrated in fi gure 12.1, which compares Greece and Ireland and provides two main messages. First, after a period of quick accumulation, Ireland rolled its debt back in the 1990s, a strong signal of dis-cipline. The Greek debt, on the other hand, has enormously increased over the whole period, yet it has been stationary over the two decades preceding the crisis. Second, why did disciplined Ireland loose control of its debt after 2007 and run afoul of markets at about the same time as Greece? The answer is that the authorities allowed for a build- up of private debts (public sector salaries were also raised in an era of plenty). When the housing price bubble blew up, the banking system collapsed and the government was compelled to rescue its banks. This illustrates an important point: fi scal discipline may be threatened in a myriad of ways that have apparently nothing to do with the regular budget. Fiscal discipline is hard to assess because unexpected spending needs may arise.

Why should governments ever be fi scally undisciplined? After all, the

3. Formally, fi scal discipline requires that the government be solvent, that is, that the trans-versality condition be upheld. This condition can never be tested because it refers to the future evolution of the public debt. Debt stationarity is one way to make the transversality condition operational.

Fig. 12.1 Debt / GDP ratios for Greece and Ireland: 1961–2009Source: Historical public debt database, IMF (2010).


budget constraint cannot be avoided. If it is ignored, it reestablishes itself through infl ation or defaults. A large literature identifi es a number of rea-sons that explain the widespread defi cit bias phenomenon.4 Although many reasons have been advanced (Wren- Lewis [2011b] offers an exhaustive list), two of them seem to dominate. The fi rst one is the tendency to push out the discipline burden to future governments or even to future generations. The second reason is the interplay of democratic processes and interest group politics. Politicians enhance their (re)election probabilities by catering to interest groups and providing public largesse, at the expense of future tax-payers. Even when there is an understanding that the process is unstable, fi nding political support to rein in defi cits may be impossible or take too long to achieve, along the lines of attrition games (Alesina and Drazen 1991).

Shifting the debt burden to future governments and spending above tax-ing level can both be seen as a manifestation of the common pool problem, which arises when the benefi ciaries of public spending (or tax advantages) ignore the externality that they impose on all other taxpayers. “Money grab” by pressure groups is an intratemporal externality, while pushing out taxes on future generations is an intertemporal externality.

The common pool problem is inherent to democratic systems unless the voters are perfectly homogeneous and care about their descendants exactly as they care about themselves. It should not come as a surprise, therefore, that the defi cit bias, the tendency of governments to run far too frequent defi cits, is such an ubiquitous phenomenon, as evidenced in table 12.1. In fact, the surprise is that some countries could be free of the bias; fi nding out how they do so is a main objective of the present chapter.

A weakness of table 12.1 is that it does not take into account the relative sizes of defi cits and surpluses, nor does it take into account the growth rate. For example, Japan only had defi cits over 68 percent of the years, but its 2011 debt, at 204 percent of GDP according to the OECD, is vastly larger than that of the Netherlands, where it stands at 78 percent of GDP while defi cits have occurred in 88 percent of the same years. A theoretically better gauge is the evolution of the debt as ratio to GDP. A country can be said to have been fi scally disciplined if this ratio has been stationary over a sustained period.5 Unfortunately, the powers of stationarity tests are famously weak. This is why table 12.2 presents the results of two opposite tests: the standard ADF test (Augmented Dickey Fuller), which asks whether a series is nonstation-ary, and the KPSS test (Kwiatkowski, Phillips, Schmidt, and Shin), which asks the opposite question—namely, whether a series is stationary. Being

4. A synthetic treatment is offered by Persson and Tabellini (2000).5. The proper concept is the transversality condition, but it cannot be observed. Requiring

that the debt be stationary is a sufficient but not necessary condition. An alternative is to look at fi scal policy reaction functions as in Bohn (1998), Galí and Perroti (2003), and Wyplosz (2006), and check whether a dynamic stability condition is satisfi ed.

Tab

le 1

2.2

Sta

tion

arit

y te

sts:

196

0–20

06

Tes

t

Nul

l

Aus

tral

ia

Aus

tria

B

elgi

um

Can

ada

AD

FN

onst

atio

nary

Rej

ect*

KP

SSSt

atio

nary

Rej

ect*

*R

ejec

t***

Rej

ect*

*R

ejec

t**

Con

clus

ion

Non

stat

iona

ry

Non

stat

iona

ry

Am

bigu

ous

N

onst

atio

nary

Fin

land

F

ranc

e

Ger

man

y

Gre

ece

Ir

elan

d

AD

FN

onst

atio

nary

KP

SSSt

atio

nary

Rej

ect*

*R

ejec

t***

Rej

ect*

**R

ejec

t***

Con

clus

ion

Non

stat

iona

ry

Non

stat

iona

ry

Non

stat

iona

ry

Non

stat

iona

ry

Am

bigu

ous

Ital

y

Japa

n

Net

herl

ands

N

ew Z

eala

nd

Nor

way

AD

FN

onst

atio

nary

KP

SSSt

atio

nary

Rej

ect*

**R

ejec

t***

Rej

ect*

Rej

ect*

*C

oncl

usio

nN

onst

atio

nary

N

onst

atio

nary

A

mbi

guou

s

Non

stat

iona

ry

Non

stat

iona

ry

Port

ugal

Sp

ain

T

urke

y

UK

U

S

AD

FN

onst

atio

nary

Rej

ect*

KP

SSSt

atio

nary

Rej

ect*

*R

ejec

t**

Rej

ect*

*R

ejec

t**

Rej

ect*

*C

oncl

usio

n

Non

stat

iona

ry

Non

stat

iona

ry

Non

stat

iona

ry

Am

bigu

ous

N

onst

atio

nary

Sou

rce:

Abb

as e

t al.

(201

0).

Not

es: T

he s

ampl

e in

clud

es a

ll co

untr

ies

for

whi

ch th

e da

ta c

over

the

peri

od 1

960

to 2

006,

aft

er fi

lling

in s

ome

mis

sing

obs

erva

tion

s w

ith

linea

r in

terp

olat

ion.

T

he A

DF

test

look

s for

a u

nit r

oot w

hile

the

KP

SS te

st c

ompu

tes a

n L

M (L

agra

nge

mul

tipl

ier)

stat

isti

c co

ncer

ning

the

vari

ance

of

a ra

ndom

wal

k co

mpo

nent

in

the

deco

mpo

siti

on o

f th

e or

igin

al s

erie

s in

to a

con

stan

t and

a r

ando

m w

alk.

***S

igni

fi can

t at t

he 1

per

cent

leve

l.**

Sign

ifi ca

nt a

t the

5 p

erce

nt le

vel.

*Sig

nifi c

ant a

t the

10

perc

ent l

evel

.


weak, both tests tend to not reject the hypothesis being tested. This is indeed what is often found when looking at countries for which comparable data on the debt- to- GDP ratios exist going back to 1960. The sample stops in 2006, in order to avoid the exceptional increases associated with the fi nancial crisis.

Table 12.2 indicates whether any of these two hypotheses is rejected by the corresponding test. Because of low test power, nonrejection of the null hypothesis is probably often misleading. More interesting, therefore, is when the test is rejected, but even so the combination of both tests can yield ambiguous results. Of the nineteen countries included in the sample, nonstationarity is only rejected twice by the ADF test (for Belgium and the United Kingdom), but in both instances, stationarity is also rejected by the KPSS test. The opposite case, when neither stationarity nor nonstationarity is rejected occurs twice, in the cases of Ireland and the Netherlands. For the remaining fi fteen countries, nonstationarity is not rejected while stationar-ity is rejected, suggesting that fi scal indiscipline is indeed the rule, with no clear exception. Interestingly, the debt crisis has (so far) only affected two of these fi fteen countries, as well as Ireland, which is classifi ed as ambiguous.

12.3 An Analysis of Policy Rules

The previous section has shown that the defi cit bias is widespread among developed countries. It has also argued that the intra and intertemporal common pool phenomena are likely to be key reasons for the defi cit bias. This bias is often cited as an argument for intervention in the political pro-cess that drives the preparation, adoption, and execution of the budget. The precise form of this intervention must be carefully tailored to achieve a second- best outcome.

In particular, if the goal is to reduce the defi cit bias via a reformed budget process, then the policy intervention needs to target the political failure that gives rise to defi cit bias, in effect internalizing the common pool external-ity as shaped by the political process. Von Hagen (2002) concludes that the solution depends on the electoral process and the form of government. In particular rules, which are time- inconsistent by construction, are likely to be dominated by institutions, as argued in Wyplosz (2005). On the other hand, political acceptability of fi scal policy institutions seems highly lim-ited, if one judges from the small number of countries that have adopted some, as described in section 12.5. Numerical rules, on the other hand, are widespread and they often have been successful, so far at least. For this reason it is worthwhile to examine fi scal rules with an open mind. This sec-tion starts with theoretical considerations, asking how fi scal rules fare on four key dimensions: the time inconsistency problem, burden sharing across generations in aging societies, capture by special interests, and government hierarchies.

502 Charles Wyplosz

12.3.1 Time Inconsistency

As a commitment device, a rule is vulnerable to time inconsistency. Indeed, there will always be instances when it is suboptimal to abide by previous commitments. Much as a rule that can be easily evaded is use-less, a rule that is strictly set stands not to be respected in some situations.6 Two main implications follow. First, it must be recognized that there exist unforeseeable events likely to break the rule. The usual solution is to write escape clauses into the rule. Escape clauses, however, cannot be fully contin-gent, because too many relevant events are unforeseeable. This opens up a major risk of circumvention. Second, some foreseeable events may warrant foreseeable adjustments to the rule without affecting its disciplinary effect. This is clearly the case for cyclical fl uctuations. While each business cycle is sui generis, in principle it is possible for countercyclical fi scal policies to be compatible with debt stability. It is also desirable because procyclical policies are particularly subject to time inconsistency.

12.3.2 Aging and Burden Shifting to Future Generations

The common pool temptation is nearly irresistible in the face of an aging population. Aging has two origins: increases in life duration and reduced fertility. The fi rst one is slow and, presumably, permanent. The second one probably is a one- off event (i.e., fertility will remain low) that will create a new steady state with a permanently smaller population. As is well known, the resulting demographic transition between two steady states creates a situation whereby the next generation, the fi rst one to be less numerous, faces a larger burden than the previous and the following ones. Part of the increase comes from higher health costs, which each generation should fi nance for itself.

But another part of the impact of the demographic transition, probably the bulk, comes from the need to pay for retirement benefi ts. In a pay- as- you- go system, each generation cares for the previous one and will be taken care of by the next one; clearly the demographic transition breaks the fair-ness of this arrangement at the expense of the generation in between. The common pool problem implies that the fi rst generation, the baby boomers, will endeavor to shift the burden to the next one. Indeed, estimates by the European Commission (2009) indicate that, given current policies, twenty of the twenty- seven EU countries will raise the annual costs of pension by more than 2 percent of GDP by 2060. The average additional cost for these countries is estimated to represent 5.8 percent of GDP.

Reducing this considerable burden shift requires that current voters approve paying more taxes, receiving less pension benefi ts, or working lon-

6. The fi nancial crisis is a good example. Many governments sincerely committed to fi scal discipline ended up reluctantly opening up large defi cits.


ger. Several countries have started to move in this direction. These moves can be seen as numerical rules that affect some specifi c elements of the budget. Even though not all countries have enacted such rules, or not sufficiently so, the fact that voters are willing to support such decisions illustrates one important advantage of rules when they are transparent: they make the externality clear and its internalization acceptable.

An alternative and intriguing interpretation is that the current genera-tion expects the next one to shift the burden as well to the following one, which would do the same, and so forth. This could be optimal intergen-erational burden smoothing. Yet another possibility is that the externality could be entirely internalized within families if baby boomers leave adequate bequests. This would be a way of reducing wealth redistribution—and per-petuating inequalities.

12.3.3 Capture

The standard common pool effect describes interest groups vying for fi nancial favors from the government, under the assumption that the corre-sponding taxes will be paid for mostly by others. In equilibrium, each interest group pays for all the favors, an externality that is the source of a defi cit bias. Obviously, the solution to the common pool problem is that the government rejects all favors, carrying out spending and transfers purely on the basis of welfare principles exactly as the mythical benevolent dictator would do. The defi cit bias arises because there is no benevolent dictator, only governments that court the support of voters and that are captured to varying degrees by interest groups. A number of empirical studies provide indirect support to this interpretation, showing that any source of heterogeneity (income inequality, political fragmentation, ethnic diversity) that enhances externali-ties ceteris paribus leads to larger defi cits.7

The response to the common pool problem is to centralize the budget pro-cess in order to centralize the externalities. A large number of solutions have been proposed and many of them have been implemented. They all aimed at improving the governance of the budget process, refl ecting the fact that the defi cit bias can be seen as a political failure, a weakness of democratic systems.8 The following sections examine and evaluate these solutions.

12.3.4 Central and Local Governments and International Aspects

Of particular interest is the relationship between central and subcentral governments. This situation opens up an additional common pool problem when local governments may reasonably expect to receive transfers from the

7. The classics are Roubini and Sachs (1989) and Alesina and Perotti (1995). For a recent contribution, see Larch (2010).

8. A survey of recent changes is IMF (2009).

504 Charles Wyplosz

central (or higher- up) governments.9 As a result, subcentral governments are subject to two mutually reinforcing sources of a defi cit bias.

A similar situation arises at the international level when a national gov-ernment receives external fi nancing, either as part of regular transfers or as bailouts. This is why the IMF only provides loans and requires that these loans be senior and fully serviced. The same applies to World Bank lending, with the exception of Individual Development Account (IDA) grants to the poorest countries.

The European monetary union can be seen as an intermediate case between a federal arrangement and a purely international agreement. There exist regular transfers, under the structural funds program and the common agricultural policy, which together absorb the bulk of a small budget (about 40 percent of a budget that represents about 1 percent of European GDP). Yet, for some countries, these transfers amount to several percentage points of GDP.10 The crucial question is whether a government can be bailed out in case of budgetary difficulties. Within the monetary union, the no- bailout clause (art. 125) was intended to be the solution to the international common pool problem. As is well known, the clause has been ignored and bailouts have been handed out to three member countries (at the time of writing). In order to compensate for the corresponding defi cit bias incentive, a strength-ened Stability and Growth Pact is currently under adoption.

12.4 Arrangements for Fiscal Policy Discipline

Arrangements to limit the defi cit bias must offset the incentives generated by the common pool problem. Whatever form they take, they must act as a constraint on the budgetary process and therefore on those who decide on the budget. The challenge is that those who have incentives to allow for a defi cit bias are asked to adopt arrangements that will eliminate or reduce these incentives. There are basically three possibilities: (1) delegation to an agent; (2) binding numerical rules; (3) a better budgetary process.

12.4.1 The Democratic Requirement

All democracies share the property that budgets are set jointly by the executive and the legislative branches. Budgets require approval by elected bodies, essentially because fi scal policy is redistributive. Taking from some to give to others is only legitimate if it is the outcome of an uncontroversial

9. See, for example, Alesina, Angeloni, and Etro (2001) and Krogstrup and Wyplosz (2010). Pisauro (2003) notes that, on top of the common pool problem, there exists a moral hazard problem created by the possibility of a bailout of local governments by the central government. While the mechanics of bailouts are indeed different from those of transfers, a bailout can be seen as part of the more general common pool problem.

10. Schuknecht, von Hagen, and Wolswijk (2008) fi nd that EU countries that are net ben-efi ciaries of transfers face a lower risk premium ceteris paribus, and that the same applies to Canadian provinces.


democratic process (“no taxation without representation”). Any constrain-ing arrangement must be democratically legitimate. How do potential solu-tions to the defi cit bias measure up to this requirement?

A fi rst solution is to delegate some aspects of fi scal policy to an unelected agent. This may well be the fi rst- best solution. Indeed, the defi cit bias can be reduced or eliminated through delegation to an agent who is not exposed to pressure by interest groups. Taking power out of the hands of elected officials in favor of bureaucrats may seem excessive, but that is exactly what is achieved with central bank independence.11 Monetary policy, however, has limited redistribution effects, mostly between borrowers and lenders as the result of interest rate changes. In the absence of infl ation, these effects are likely to be reversed along the cycle. For this reason, delegation of mon-etary policy is different from delegation of budgetary responsibilities. Even though delegation could be limited to just setting the budget balance, which has limited redistributive effects, it is not observed in practice. Delegation to nonelected officials appears to be difficult, if not impossible.12

Numerical rules are more acceptable because they are self- imposed by the very elected officials that they are designed to bind. In fact, rules are like laws that restrict freedom for the common good. Governments and parliaments routinely operate under such laws, many of which are inscribed in constitu-tions. Naturally, elected officials wish to avoid limits on their own freedom of action. Since laws can be amended or even repudiated, the challenge is not just to have such limits adopted, but that they remain in place when they become binding. There are several instances of potentially useful rules that were rescinded when they were needed most. This is the case of Europe’s Stability and Growth Pact or of former Chancellor of the Exchequer Gor-don Brown’s self- imposed fi scal responsibility principle. Another example is the US Gramm- Rudman- Hollings defi cit reduction law of 1985; the annual defi cit target was raised in 1987 when it proved politically difficult to meet.

The last solution avoids any formal straitjacket. Instead, it aims at encour-aging policymakers to internalize among themselves the externalities that give rise to the defi cit bias. Various solutions have been suggested. Haller-berg, Strauch, and von Hagen (2009) distinguish between two categories: delegation and contract. Under delegation, some power is vested with a player, for example the fi nance minister, whose role is to achieve fi scal disci-pline. Under contract, the political parties that join in a governing coalition agree upon fi scal discipline. Hallerberg and collegues note that the success of any arrangement depends on whether it is compatible with the politi-cal backstage. For example, coalition governments are unlikely to devolve strong power to a fi nance minister who necessarily represents one party

11. In noneconomic matters, other references include a Supreme Court of Justice and Con-stitutional Courts.

12. Wren- Lewis (2011a) makes a similar point.

506 Charles Wyplosz

and its supporting interest groups. Another type of attempt at improving the budget process is to empower somewhat the administration, seen as less likely to bend to pressure groups. In this case the bureaucrats do not make decisions, but sort out and shape the options that are presented to policy-makers.

12.4.2 Types of Rules

Existing numerical rules vary from one country to another. They fre-quently stipulate upper limits on the budget balance, the debt, spending, or they set lower limits on tax revenues. Quite often, several of these limits are combined. Some rules apply year by year, others defi ne limits over several years, sometimes over the entire business cycle or over the duration of a government. Some are national, with subcentral government components. The European Monetary Union’s Stability and Growth Pact is a suprana-tional rule.13 According to IMF (2009), eighty countries have adopted some rule or another.

In theory, if they are well- designed and implemented, fi scal rules can elim-inate the defi cit bias. In practice, however, rules are often disappointing. A fi rst difficulty harks back to the old debate on rules versus discretion, and the time inconsistency problem. Because rules can never be fully contingent, situations may arise that would make any rule very costly to respect. The fi nancial crisis, which has led to cumulated debt increases of some 30 percent of GDP in many developed economies, is a case in point.

For this reason, any rule must be fl exible enough to accommodate unfore-seeable contingencies. As a consequence, the precise design of rules becomes of the utmost importance. The presumption is that rules should be simple enough to be understood by policymakers and citizens alike, but fl exibil-ity is bound to come at the expense of simplicity. Consider, for example, strict balanced budget rules, which imply procyclical fi scal policies. In order to avoid procyclicality, and possibly to encourage countercyclicality, the rule must target the budget over a whole cycle or be stated in terms of a cyclically adjusted measure. Since business cycles cannot be predicted, an over- the- cycle rule can lose much of its meaning, although solutions have been proposed, as discussed later. Cyclical correction is more art than sci-ence and is not easily comprehended by the public at large, which opens the door to manipulations and, quite possibly, to an eventual repeal.

More generally, rules can always be manipulated. Crucially, since they are forward- looking, budgets are constructed on the basis of assumptions about economic and fi nancial conditions over the next fi scal year. This gives much leeway to the government, often enough to loosen the rule. Independent ex

13. Both African monetary unions and the East Caribbean monetary union also operate supranational fi scal rules.


post evaluations of outcomes is a solution, but it involves judgment and, anyway, evaluations will always come too late.14

Furthermore, the democratic requirement articulated in the previous sec-tion implies that the politics of fi scal rules is not encouraging. A government subject to the defi cit bias can fairly easily convince its public opinion that today’s circumstances are special and that technocratic arrangements should not stand in the way of serving people’s interests. In addition, rules only work if noncompliance is sanctioned with a high degree of certainty. In democ-racies, however, voters seldom sanction governments that fail to deliver on their promises, unless they violate the law. Fiscal discipline rarely defi nes elec-tion outcomes—otherwise we would not observe the frequent lapses docu-mented in table 12.1. Legal sanctions require that the law be written in precise enough terms to face powerful challenges, which runs against the simplicity principle. Worse, laws can always be changed at the worst time if they are time- inconsistent. Debrun (2011) provides an example where a rule is time- inconsistent for the currently elected government but not for the public at large.

This all implies that fi scal rules are unlikely to be a panacea. This is indeed what the evidence suggests (see, e.g., Kopits 2001; Guichard et al. 2007; and IMF 2009). On the other hand, Debrun and Kumar (2007) provide panel- data estimates of the impact of fi scal rules on the primary budget balance and fi nd that rules can be effective. Yet, the evidence is not overwhelming. This may be because rules differ too widely in their details for valid cross- section comparisons. It may also refl ect that the unobserved political and institutional context matters. Another interpretation, suggested by Debrun and Kumar (2007), is that the evidence may suffer from reverse causality, namely that disciplined governments may wish to adopt rules as a way of cementing and signaling their determination.

12.4.3 Rules versus Institutions

The limits of rules may be seen as making fi scal institutions an attrac-tive alternative. Defi ned broadly to include nonnumerical rules, fi scal insti-tutions encompass a wide variety of arrangements. Examples include the (possibly partial) delegation of the budget process to an independent body, intragovernmental agreements, multiyear programming, and codes of good behavior. The advantage of institutions is that they are less likely to be time- inconsistent, if only because they do not require specifying all contingen-cies that may arise and what to do when they arise. Institutions instead can have a mandate (to establish fi scal discipline in the long run) and they may involve people who can think. If these people are sufficiently independent and competent, mandates should do better than rules.

14. A solution is to require that outruns be compensated (see the Swiss Debt Brake). Even so, massive crises like the current one cannot be dealt with easily.

508 Charles Wyplosz

This means that institutions can work but only if they are well- designed. What are the conditions for an effective design? Hallerberg, Strauch, and von Hagen (2009) convincingly argue that there is no single answer to that question. In order to work well, institutions must be adapted to the political institutions, including electoral systems, types of governments, and degree of ideological differences in society. They provide evidence that well- adapted institutions work.

Fiscal policy councils are a particular case of fi scal institutions. Their remits can vary all the way from just “telling the truth” to actually decid-ing on the budget balance. In the earlier case, the council is intended to be an official watchdog, one that benefi ts from official recognition and pro-vides unbiased views. For example, the council may be formally consulted to evaluate the government forecasts of spending and income, which are based on forecasts of variables such as GDP growth, infl ation, the interest and exchange rates, and so forth. Its own forecasts may be for mandatory use in budget planning or simply produced for advisory purposes.

Up to now, there does not exist any fi scal policy council with the power to decide on the budget balance, probably because it would clash with the democratic requirement. On the other hand, councils with formal advisory roles have become frequent in recent years. Calmfors and Wren- Lewis (2011) list eleven independent councils with advisory or nonbinding control roles in developed countries.15 Debrun, Hauner, and Kumar (2009) report that similar councils have been set up in other developed or emerging- market countries (Japan, Chile, Indonesia, Jordan, Korea, and Mexico). Several international institutions, including the IMF, the OECD, and the European Commission, have recommended the establishment of such councils.

A growing empirical literature has begun to evaluate the design and effec-tiveness of fi scal policy councils. Country studies (e.g., Calmfors 2010; IMF 2005; and Debrun, Hauner, and Kumar 2009) suggest that, in order to have a detectable disciplinary impact, fi scal councils must be allowed to make normative, quantifi ed statements. Yet it also appears that the political costs for a government not to heed such advice are very limited.16

A fair conclusion is that advisory fi scal policy councils have made a tan-gible contribution to fi scal discipline in countries where policymakers have shown a willingness to listen to them, which is why they created them in the fi rst place—a case of reverse causality. Elsewhere, the councils provide

15. These committees are in Austria (Government Debt Committee established in 1997); Bel-gium (High Council of Finance, 1989); Canada (Parliamentary Budget Office, 2008); Denmark (Economic Council, 1962); Germany (Council of Economic Experts, 1962); Hungary (Fiscal Council, 2008); Netherlands (Central Planning Bureau, 1947); Slovenia (Fiscal Council, 2010); Sweden (Fiscal Policy Council, 2007); the United Kingdom (Office for Budget Responsibility, 2010); and the United States (Congressional Budget Office, 1975).

16. A good example is Hungary. A fi scal policy committee was set up in 2008 but was effec-tively dismantled a couple of years later when a new government came to power.


useful analyses and viewpoints but their recommendations are frequently ignored. Like fi scal rules, advisory fi scal policy councils are not a panacea.

12.4.4 Rules and Institutions

There is a tendency to consider rules and institutions as substitutes. Yet, the limitations of each approach suggest that combining them may help. Because they can never be adequately contingent, rules are too rigid and therefore time- inconsistent; they simply cannot be respected in some situ-ations. Fiscal institutions may be seen as too open- ended and therefore time- inconsistent in the sense that they may be too fl exible in the face of unforeseen events. But fi scal institutions that apply and interpret not fully contingent rules are promising. Like a supreme court that applies and inter-prets laws voted by the parliament, or like a central bank that follows a fl ex-ible infl ation- targeting rule, fi scal institutions can appeal to a rule to guide and justify their actions. Rules can be deviated from when needed without losing their credibility if an independent and competent institution autho-rizes such deviations. The examples that follow illustrate the benefi ts from combining rules and institutions.

12.5 Historical Experiences

This section presents a few examples of fi scal arrangements from devel-oped and emerging market countries. While most arrangements have been introduced rather recently, other countries have long sought to buttress fi scal discipline.

12.5.1 Central Government Rules and Institutions

The Netherlands

Following years of defi cits, the Netherlands started to build original arrangements in the early 1980s. These arrangements have been gradually refi ned. As fi gure 12.2 shows, the Netherlands has performed better than the other European countries. Its public debt has been reduced from 1993 onward, until the onset of the fi nancial and economic crisis. Even though part of the measured (gross) debt reduction is related to asset sales and to natural gas revenues, the prima facie evidence—based on both timing and the fact that these actions have not been squandered—is that the Nether-lands has become fi scally disciplined following the adoption of its new fi scal policy regime.

The Dutch arrangement combines rules and institutions.17 The key rule is a path for the budget ceiling, determined in constant euro and set for the duration of each parliament. The institutional setup is quite elaborate and

17. For a detailed presentation, see Bos (2007).

510 Charles Wyplosz

rests on an explicit agreement among coalition parties before they take office and is valid until the next election. Two bodies are involved: the CPB (Neth-erlands Bureau for Economic Analysis) and the SER (Social and Economic Council of the Netherlands).

The CPB is a technical agency that makes economic and budget forecasts. The originality of the Dutch approach is that it evaluates the budgetary implications of political party programs before elections. After the election, the parties that form a coalition work out a binding medium- term program for the duration of the legislature. This program is evaluated by the CPB, taking into account its own macroeconomic projections. Then, as annual budgets are being prepared (preparations start two years ahead), the ministry of fi nance relies exclusively on CPB forecasts and medium- term projections. As a highly respected neutral agency, the CPB thus takes the responsibility for macroeconomic and budget forecasts out of political hands.

The SER is a tripartite advisory body composed of one- third employers, one- third trade unions, and one- third experts, including the CPB. It deals with the social security system and with pensions (and wage negotiations). Its recommendations are not binding on the government but infl uential when it can reach an agreement.

A key element is the required adoption of medium- term spending ceil-ings (for each ministry) at the beginning of the legislature. These ceilings are based on a detailed list of policy measures agreed upon in the coalition

Fig. 12.2 Gross public debt in the Netherlands and the Eurozone (percent of GDP)Source: AMECO, European Commission.


agreement. The agreement itself is built around explicit defi cit and debt targets for the end of the legislature. There is no standard—and arbitrary—mandatory target numbers but, given that governments always involve sev-eral parties, the practice has a clear moderating impact.18 There is a little bit of fl exibility (1 percent) for shifting spending from one year to the next but “growth bonuses” are not usable later. The result is a tendency for procycli-cal policies. This aspect may be of limited importance for a small and very open economy. Taxes become the main macroeconomic instrument, under the scrutiny of the CPB, which evaluates debt sustainability.

The effectiveness of the Dutch arrangement can be traced to three ele-ments. First, the combination of rules and institutions. Second, the fact that the numerical rule is not tied to a particular, arbitrary number but left to publicly visible negotiations among coalition partners. Third, they are well- adapted to the Dutch political system. As argued by Hallerberg, Strauch, and von Hagen (2009), legislative contracts work well in multiparty government coalitions.

Switzerland and Germany

The constitutional rule adopted by Germany in 2009 is due to be fully implemented in 2016.19 It is closely patterned after the Swiss “debt brake” that was adopted and written into the constitution in 2000 and came into force in 2003. Figure 12.3 shows that the rule has reversed the familiar trend of debt build- up, a reversal particularly spectacular as it has continued throughout the crisis.

The debt brake is a rule, with an escape clause that involves the parlia-ment. In the Swiss version, the rule specifi es that the overall federal budget must be balanced over the cycle. This is achieved as follows. Any imbalance, positive or negative, is credited into a control account. If the cumulated amount is negative, it must be brought to balance “over the next few years.” No requirement applies when the cumulated amount is positive. This clever arrangement implies that, over time and at the government’s discretion, defi cits must be compensated for by surpluses. The stipulation is fl exible enough not to put the government in a procyclical straitjacket. It can lead to prolonged slippages, however. In a country very sensitive to the rule of law, like Switzerland, such slippages are highly unlikely.20 During the crisis, the debt brake fi gured prominently in policy debates and quite clearly shaped the policy response.

The arrangement includes an escape clause. In case of exceptional circum-

18. Expenditures had crept up to almost 60 percent of GDP by the early 1980s. They went down to less than 45 percent before the crisis.

19. It was initially planned to implement it in 2011 but this date has been pushed back because of the crisis. Interim arrangements start in 2011, however. A variant will apply to the Laender as of 2020.

20. It remains to be seen whether it will be as effective in Germany once it is implemented.

512 Charles Wyplosz

stances (deep recession, natural disasters, and the like) the implied spending ceiling can be raised but this requires a qualifi ed majority (three- fi fths) in both chambers. An amendment stipulates that any such slippage must be added to the control account, which means that, even if exceptional, bygones shall not be bygones.

The success of the Swiss debt brake (so far) derives from the simplicity of the rule. As previously noted, it remains to be seen what would happen should a negative balance in the control account not be corrected “over the next few years.” Presumably, the case could be sent to the Higher Court. That the debt has continued to decline throughout the crisis is surprising since defi cits are expected in bad years, even without invoking exceptional cir-cumstances (which would not have applied since the GDP declined by only 2 percent in 2009). One reason is that Switzerland is a small open economy closely integrated with the European Union so that it benefi tted from fi scal expansions implemented elsewhere. Another reason is that the debt brake only concerns the federal government, which spends about one- third of total public outlays. The rest of the spending is carried by the cantons (about 40 percent of the total), municipalities, and the social security system. Many, but not all, cantons have adopted since the early 1980s—and some much ear-lier—various forms of budget rules. As a result, with one exception,21 their debt levels are low. The same applies to municipalities in a similar fashion.

Fig. 12.3 Swiss public debt (percent of GDP)Source: Economic Outlook, OECD.

21. The exception is the Geneva canton, which has no budget rule.


Chile

Chile is an early fi scal rule adopter. Adoption came at the end of a long period during which the public debt was reduced from 165 percent in 1985 to 20 percent of GDP by 2000. The intention was to solidify and codify the emerging fi scal discipline tradition.22 Maybe because of its pioneering aspect, the rule is technically complex (for noneconomists) and not quite complete.23 Initially introduced informally, the rule was written into law in 2006 (Fiscal Responsibility Law).

The rule requires that the cyclically adjusted primary budget be in surplus. The target was 1 percent from 2001 until 2008, when it was reduced to 0.5 per-cent, and further to 0 percent in 2009 to allow for a countercyclical response to the global crisis. In practice, the procedure is to estimate cyclically adjusted government revenues and to then derive total maximum spending. There is no escape clause, but the target can be changed, as already noted. While the budget must be ex ante in conformity with the rule, there is no sanction when the realized budget differs, presumably because underlying assump-tions proved to be too optimistic, which happened twice, in 2002 and 2009.

The calculation of the ex ante structural revenues is therefore the lynchpin of the rule. Beyond the usual complexity of taking into account cyclical effects, Chile’s rule is highly sensitive to copper price fl uctuations because tax revenues from copper production can represent a quarter of total public income. Given the volatility of copper prices, it is essential but highly chal-lenging to correct ex ante for these fl uctuations. This is where an institution is needed to operate the rule. A committee of independent experts is in charge of providing the government with assumptions regarding GDP and the long- run price of copper. It follows that forecast errors concerning the GDP cannot be the result of government’s manipulation. It also follows that in any given year the actual price of copper is likely to differ from the long- run estimate. It is expected that the independent experts are not biased, so that forecast errors must cancel out over time, thus imparting no bias to the rule.

The surplus rule implies that eventually, the government must be a net creditor, which occurred in 2005, as fi gure 12.4 shows. This was an objective of the rule, in fact. It was recognized that copper resources would eventually disappear and that Chile needed to build up a welfare system. As a result, Chile operates two Sovereign Wealth Funds.

The success of the Chilean rule may be surprising given its incomplete-ness. There is no limit to the possibility of changing the target, nor even any standard procedure to do so, and there is no sanction for not achieving the target. Part of the reason for the success is the existence of an expert group, which is involved in the budgetary process, as it produces cyclically adjusted fi gures, an arrangement that is remindful of the Dutch case. Another reason

22. This is a clear case of reverse causality in the sense of Debrun and Kumar (2009).23. Dabán (2011) offers a recent review.

514 Charles Wyplosz

is the transparency of the process. The calculations of the cyclically adjusted budget, which have been refi ned over time, are presented and explained in great detail to the broad public. The two recent target changes, for instance, have been carefully and candidly explained.

Equally important is the relationship between government and parlia-ment. The power to set the budget is entirely in the hands of the president. The parliament is not allowed to reduce taxes or raise spending. This fi ts well with the delegation model of Hallerberg, Strauch, and von Hagen (2009) since the presidential system delivers de facto a single party majority.

Britain’s Office for Budget Responsbility

In 1997, the British Chancellor of the Exchequer adopted two fi scal rules: (1) the budget defi cit may only fi nance public investment (a golden rule); (2) the debt- to- GDP ratio may not exceed 40 percent. The rule was to be monitored by the chancellor himself, based on forecasts by HM Treasury. The Treasury was also requested to produce long- run forecasts (forty years) to gauge long- run sustainability. As fi gure 12.5 shows, the rule was met for a few years but then slippage set in after 2002. The idea that a public com-mitment with no enforcement mechanism would work was disproved. This has led the newly elected British government to pass in 2011 the Budget Responsibility and National Audit Act.

The new act combines a set of rules and an independent fi scal policy com-mittee.24 The rules are similar to the previous ones: the cyclically adjusted

Fig. 12.4 Chile’s central government (percent of GDP)Source: Dabán (2011).

24. For details, see Wren- Lewis (2011b). This arrangement bears similarity to the Swedish independent council, described in Calmfors (2010).


budget must be balanced—or in surplus—over the following fi ve years, the debt- to- GDP ratio must be declining by the end of the legislature, and long- run forecasts must show that public fi nances are sustainable. The main difference is the replacement of the golden rule—a shaky concept since deciding what is public investment is open to manipulation—by the balance requirement over a rolling fi ve- year horizon, which is essentially calling for balanced budgets, on average, over time.

The real innovation is the creation of the Office for Budget Responsibil-ity (OBR). This independent body takes over the forecasting tasks so far carried out by the Treasury. This arrangement is a clear response to the fact that many of the slippages over the 2000s were predicted by optimistic forecasts regarding both macroeconomic variables and the budget fi gures. A Budgetary Responsibility Committee that includes fi ve members runs the OBR. The fi rst committee includes widely respected economists. The staff is small (fi fteen persons), under the assumption that the OBR has full access to the Treasury resources.

Much like the US Congressional Budget Office (CBO), the OBR is restricted to produce forecasts only on the basis of government policy. This implies that the OBR cannot look at “what if” questions; that is, to make policy suggestions. On the other hand, it has the monopoly of official fore-casts and policy evaluations. Importantly, the OBR has only an advisory role in the sense that the chancellor can carry out any policy that he wishes, but

Fig. 12.5 Public debt in Great BritainSource: AMECO, European Commission.

516 Charles Wyplosz

under the constraint that he cannot challenge the OBR’s macroeconomic and budgetary implications since the Treasury has given up the possibility to use its technical expertise to that effect. Since the chancellor is also bound by the rules, the room for undisciplined behavior is narrow, but it still exists.

It will obviously take time to evaluate whether this arrangement is effec-tive. It rests on delegation of fi scal discipline to the chancellor; Hallerberg, Strauch, and von Hagen (2009) argue that this is the correct arrangement for single- majority governments, which often occurs in the United King-dom but not currently. Delegation also characterized the 1997 arrangement, when a single- majority government was in place, but it failed when the chan-cellor single- handedly decided to overlook his self- imposed rule. The new arrangement combines a reasonable set of rules with an institution that is intended to act as a whistle- blower. This is in line with the view by Jonung and Larch (2006), which considers that delegating forecasting to an inde-pendent agency is a key ingredient to achieve fi scal discipline. Interestingly, in the autumn of 2011, the OBR has warned that the fi scal retrenchment efforts were too strong at a time when a recession is likely. The chancellor promptly announced a set of expansionary policies.25

12.5.2 Subcentral Government Rules and Institutions

The additional common pool problem that arises in federal states requires special treatment to avoid that subcentral government indiscipline under-mines adequate policies carried out at the central level. A textbook example of this problem is Argentina in the late 1990s.

Belgium

In the late 1980s, following years of internal strife and rapidly rising public debt, Belgium shifted from a centralized to a federal state. It created a com-plex overlapping system of regions and communities to which many cen-tral functions were transferred. By the mid- 1990s, it set up an arrangement designed to establish overall budget discipline. The key characteristic was the adoption of institutions, without any numerical rule. The main element was a reform of the High Finance Council (HFC), entrusting this body with the task of monitoring regional governments and of suggesting budget bal-ances for the various levels of government, with the objective of achieving debt sustainability. These suggestions shape formal agreements (called bud-getary conventions) between the federal and the subcentral governments. These agreements are coherent with the Stability and Growth Pact’s stability programs.26

The HFC is composed of high- level experts from ministries, the National

25. A similar situation arose in Sweden earlier in the year.26. Macroeconomic forecasts are proposed by the National Accounts Institute, which pools

resources from various institutions.


Bank, the Federal Planning Bureau, and academia. It is therefore not an independent body, rather a coalition of the willing whose task is to fi nd acceptable proposals in a system that is not politically harmonious.

The arrangement has worked well, as total public debt went down from 140 percent of GDP in 1993 to 88 percent in 2007. Since the crisis started, however, defi cits have grown again and the overall debt is expected to reach 104 percent in 2011. Figure 12.6 shows that the relapse is mostly occurring at the federal level, even though slippages occur across the board. This refl ects the working of the automatic stabilizers but also the need for the federal authorities to recapitalize banks to the tune of 6 percent of GDP. More wor-risome is the recent deterioration. It may be the consequence of an ongoing and protracted political crisis that has left the country for months running with interim governments. It may also be a consequence of the common pool inherent to the overlapping of the federal system. This system gives rise to a severe vertical imbalance, whereby spending powers are extensively devolved, whereas tax policy and collection remain in federal hands. This leads to massive implicit transfers to the Walloon and undermines account-ability at the regional level.

In a country with multiple and partly overlapping layers of governments,

Fig. 12.6 Budget balances of various Belgian authorities (€ mn.)Source: Belgostat.

518 Charles Wyplosz

each one itself a coalition, the arrangement has emphasized contracts among parties and among governments. The HFC is generally credited for having played an important role, outlining consensus solutions that respect fi scal discipline.27 However, Coene and Langenus (2011) argue that the HFC had infl uence when government’s preferences and HFC’s mandate were well aligned, yet another case of reverse causality. The absence of any numerical rule did not seem to have hampered the arrangement. Similarly, the fact that the HFC is not independent but instead brings together the traditional “cul-prits” of the defi cit bias, points to the advantage of well- structured negotia-tions. In brief, the Belgian experience conforms to the view, developed by Hallerberg, Strauch, and von Hagen (2009), that arrangements well- adapted to the political situation may deliver fi scal discipline, in contrast to the view, presented here, that favors combining numerical rules and independent bodies.

Canada and the United States

Both countries share the absence of any constraining rule imposed by the federal on to subfederal entities, compensated by the adoption by most subfederal governments of self- imposed rules. In both countries, federal spending about matches in size subcentral spending with signifi cant trans-fers from the federal level, some of which are conditional on policies in place.

In Canada, a spending limit was adopted at the federal level from 1992 to 1996. During this period six provinces adopted balanced budget rules that remain in effect. These rules differ from one province to the other (see Tellier and Imbeau 2004). Figure 12.7 shows that these changes came at a time when fi scal discipline was seriously under threat. The fi gure also shows that, since then, both levels of governments have stabilized their debts. The overall public debt level peaked at 102 percent of GDP in 1996 and declined to 67 percent in 2007. Not all provinces, however, have been fi scally disciplined: the aggregate performance conceals the combination of large surpluses in resource- rich provinces and sizable defi cits elsewhere, notably in Ontario, the largest province.

In the United States too, all states but one have adopted various forms of balanced budget rules. These rules started in the middle of the nineteenth century and have been evolving since.28 The federal government is subject to a nominal debt ceiling that is regularly raised by Congress, a procedure that can be confl ictual when the president and the Congress majority are from different parties. In 1985, Congress passed the Gramm- Rudman- Hollings Act, which introduced the notion of budget balance paths, aiming at a bal-ance in 1991. The Act was modifi ed—including after being declared uncon-

27. For a recent appraisal, see Aneja et al. (2011).28. Interestingly, Tejedo (2007) notes that these rules originated when US states faced sudden

debt increases following bailouts of private banks. For a recent description of the situation, see NCSL (2010).


stitutional by the Supreme Court—until it was abandoned in 1990, replaced by the Budget Enforcement Act, which introduced a rule of unchanged defi cits at the margin. As Figure 7 shows, this act did not prevent a serious deterioration of the federal budget until it was left to expire in 2002. This is a good example of a rule that failed, because it was too tight and because the polity did not support it.

Figure 12.7 also shows that the federal budget has been strongly coun-tercyclical while state budgets were very moderately countercyclical. This observation confi rms that the state- level rules are binding while the federal

Fig. 12.7 Federal and subfederal budget balances in Canada and the United States (percent of GDP)Source: Office of Management and Budget and Statistics Canada

520 Charles Wyplosz

arrangements are not. It is likely that state governments have little incentive to deviate from their rules because the fi scal multiplier is bound to be low given the integration of the US markets, while the federal government can expect its own fi scal policy to be more effective. Countercyclical transfers from the central level also reduce the pressure on state governments.

One interpretation of the contrasted experiences of Canada and the United States is that the common pool problem among subcentral govern-ments is relatively weak. This may be due to the small multipliers, which matches Musgrave’s view that countercyclical policy is better contracted out to the central government. Still, it remains to be understood why the com-mon pool problem has been relatively benign at the subfederal level (Cali-fornia being a counterexample). The existence of rules is one interpretation. Yet another one is the role of market- imposed discipline. The evidence is that fi nancial markets impose risk premia on government bonds that is related to debt size and also to budgetary arrangements.29 This provides both a gauge and an incentive for state governments to be fi scally disciplined.

The Euro Area

The euro area is a very special case that combines national sovereignty, as far as fi scal policy is concerned, with some elements of federalism, the common currency. Further complicating the situation is the Stability and Growth Pact, which is designed to infl uence national fi scal policies but cannot override national sovereignty. The sovereign debt crisis has further erased the separation between a common currency and independent fi scal policies after the de facto abandonment of a treaty- level no- bailout rule that prevented governments and the central bank from helping out countries unable to borrow to serve their debts and pay for their defi cits.

The European Commission can be seen as the equivalent of a central government but its budget, about 1 percent of the EU GDP, is far too small to play any macroeconomic role. In contrast to real federations, any countercyclical policy must therefore rest with the “subfederal” member states. In addition, the commission’s budget must be ex ante in equilibrium. Furthermore, both the size and the use of the commission budget must be approved by the European Council that brings together the heads of state and government. In a nutshell, the commission budget is subject to a rigor-ous equilibrium rule and is under the control of member states.

The Stability and Growth Pact can be seen as a “federal” rule inasmuch as it commits—in theory—the member countries. In practice, however, the pact has not been effective. Over the fi rst thirteen years of existence of the euro, the twelve initial member countries together have satisfi ed the 3 percent

29. For the United States, see Poterba and Rueben (1999) and for Canada see Booth, Georgopoulos, and Hejazi (2007). For the role of budgetary institutions, see Schuknecht, von Hagen, and Wolswijk (2008).


budget defi cit limit only 60 percent of the time. The number of years when the limit has been exceeded is shown for each of these countries in table 12.3. Even if the rule may have had a moderating infl uence, the frequency of its violation is so high that its credibility is low, at best.30

On the other hand, the member countries have adopted a variety of bud-getary arrangements, which are presented in European Commission (2006), although this tally is now partly outdated. The number of these arrange-ments is impressive but the devil hides in the details and most of these arrangements are effectively soft. Furthermore, few of them are directly tied to the Stability and Growth Pact, a discrepancy that is to be reduced with the adoption in 2011 of the “European Semester.” This new procedure aims at tightening the macroeconomic assumptions to be used subsequently in national budget laws by allowing the commission to evaluate them early on. Yet, as with the Stability and Growth Pact, fi nal authority rests with national authorities so that the effectiveness of this new procedure remains to be seen.

The lack of discipline at the national level (with notable exceptions, see table 12.1) and the inherent inability of the center to promote debt sus-tainability has been well- known for some time. The sovereign debt crisis illustrates the dangers of this situation. The loss of the no- bailout rule (see section 12.3.4) has considerably weakened an already feeble arrange-ment. Current efforts focus on strengthening and widening the Stability and Growth Pact, which has repeatedly failed.

The euro area experience shows that numerical rules—the defi cit and debt limits of the Stability and Growth Pact—that are neither supported

Table 12.3 Number of years with budget defi cit in excess of 3 percent of GDP: 1999–2011

Austria Belgium Finland France

4 3 0 7

Germany Greece Ireland Italy

5 13 4 8

Luxembourg Netherlands Portugal Spain

0 4 10 4

Source: AMECO, European Commission.Note: All tests apply the Newey- West procedure to determine the bandwidth.***Signifi cant at the 1 percent level.**Signifi cant at the 5 percent level.*Signifi cant at the 10 percent level.

30. An escape clause allows the limit to be exceeded, but an escape clause must be exceptional for the rule to be meaningful.

522 Charles Wyplosz

by hard legislation nor endorsed by the political system are not sufficient to deliver fi scal discipline. One could see the European Commission as a sort of advisory fi scal council dedicated to establishing discipline. The lesson here is that, in a politically heterogeneous situation (in contrast with the Dutch case) the council must be politically independent, which the commission is not, as its members are known to receive “informal” instructions from their governments, and as the commission has its own vast and complex agenda. Alternatively, following the Belgian example, the council must be ad hoc, with a narrow fi scal discipline agenda, and include policymakers and non-policymakers with the explicit objective of seeking agreements along the contract principle of Hallerberg, Strauch, and von Hagen (2009).

12.5.3 When Do Rules Actually Work?

Fiscal rules are rather brutal instruments and they suffer from time incon-sistency. When they bind, policymakers are likely to try and evade them. Policymakers can look for loopholes, they can just ignore the rule, or they can change them. Rules can be made less brutal through the adoption of escape clauses, but then they are unlikely to be effective. The nature of the common pool problem is that policymakers often fi nd it politically reward-ing not to be fi scally disciplined. As long as the political costs of ignoring the requirements of fi scal discipline outweigh the political benefi ts of letting the public debt grow, a government will choose the latter over the former. Rules are useful if they signifi cantly raise the political costs of fi scal indis-cipline. The implication is that, to be effective, rules must be embedded in institutional arrangements.

The range of possible institutional arrangements is vast. They must be well- adapted to each country’s political system. Electoral laws shape the type of government: single- party, coalition, or minority. As argued by von Hagen and Harden (1994), single- party governments make it possible to delegate power while the other forms of governance rely instead on contracts. Eth-nic or ideological diversity results in worse common pool effects and calls for “wise men” arrangements, whereby independent bodies can achieve a reputation for truth telling. In all cases, transparencies, not just concerning accounting accuracy but also about the budgetary assumptions and their implications, are required to maintain the integrity of the rules.

Many countries operate several rules such as defi cit and debt ceilings, spending ceilings, and tax fl oors. The multiplicity of rules offers opportuni-ties for the government to escape them: rule arbitrage allows a government to argue that it meets some rules and cannot possibly meet all of them, and to pick those that are less constraining at any moment of time. Likewise, complex rules are a source of opacity that makes them understandable only to a small group of experts, chiefl y government officials and interest groups that can capture them.

Because defi cits and debts are endogenous—to cyclical conditions, to


occasional fi nancial turmoil, and other one- off events—the rules can never be completely simple and mechanical. If the goal is to achieve sustained defi cit reduction, then judgment cannot be dispensed with and indepen-dent bodies such as fi scal councils can play an important role. These bodies appear to be most effective in reducing defi cit bias when they validate the adequacy of budget assumptions and the ensuing calculations, and when they are given a formal and transparent advisory role.

12.6 Conclusions

In many developed countries, the fi nancial crisis has merely added a layer of public debt to already impressive stockpiles. In these countries, fi scal policy has lacked discipline for several decades, for well- known reasons. The power of interest groups in most democracies creates externalities that lead to a defi cit bias. The current public debt crisis in several nations has led to calls to reverse the long- standing defi cit bias, which seems unlikely to occur without a change in the budgetary process, specifi cally designed to reduce the bias.

Fiscal rules have attracted increasing attention and many countries have adopted some rules. Evidence is now being accumulated on what rules can and cannot do. In line with these results, this chapter argues that rules are neither necessary nor sufficient to achieve fi scal discipline. Yet they can and do help. The chapter argues that rules can usually be dismissed all too easily when they clash with broad political objectives: discipline is pushed back to tomorrow just when it is needed most. This means that lasting rules can-not be too tight, but they become useless if they are too soft. The fi ne line between tight and soft is extremely hard to determine and may change as circumstances change. This difficulty can be alleviated through the setting- up of institutions that support the rule.

In a symmetric fashion, fi scal institutions are neither necessary nor suffi-cient to achieve fi scal discipline, but they help. Here again we face a deli-cate balance. Institutions must bind the policymakers without violating the democratic requirement that elected officials have the power to decide on budgets. This argues against assigning wide discretionary powers to fi scal institutions, but it is fully compatible with giving them either the authority to apply legal rules or to act as official watchdogs.

References

Abbas, S. Ali, Nazim Belhocine, Asmaa ElGanainy, and Mark Horton. 2010. “A Historical Public Debt Database.” IMF Working Paper 10 / 245. Washington, DC: International Monetary Fund.

524 Charles Wyplosz

Alesina, Alberto, Ignazio Angeloni, and Federico Etro. 2001. “Institutional Rules for Federations.” NBER Working Paper no. 8646. Cambridge, MA: National Bureau of Economic Research, December.

Alesina, Alberto, and Allan Drazen. 1991. “Why Are Stabilizations Delayed?” Amer-ican Economic Review 81:1170–88.

Alesina, Alberto, and Roberto Perotti. 1995. “Fiscal Expansions and Adjustments in OCED Countries.” Economic Policy 10:207–48.

Aneja, Sumit, Kevin Cheng, Yingbin Xiao, and Irina Yakadina. 2011. Belgium: Selected Issues. Washington, DC: International Monetary Fund.

Bohn, Henning. 1998. “The Behavior of US Public Debt and Defi cits.” The Quarterly Journal of Economics 113:949–63.

Booth, Laurence, George Georgopoulos, and Walid Hejazi. 2007. “What Drives Provincial- Canada Yield Spreads?” Canadian Journal of Economics 40:1008–32.

Bos, Frits. 2007. “The Dutch Fiscal Framework History, Current Practice and the Role of the CPB.” CPB Document no. 150. CPB (Central Planbureau) Nether-lands Bureau for Economic Analysis.

Calmfors, Lars. 2010. “The Swedish Fiscal Policy Council—Experiences and Les-sons.” Unpublished Paper. Swedish Fiscal Policy Council and Stockholm University.

Calmfors, Lars, and Simon Wren- Lewis. 2011. “What Should Fiscal Councils Do?” CESifo (Group Munich–Ifo Institute, Center for Economic Studies) Working Paper Series no. 3382.

Coene, Luc, and Geert Langenus. 2011. Promoting Fiscal Discipline in a Federal Country: The Mixed Track Record of Belgium’s High Council of Finance. National Bank of Belgium.

Dabán, Terersa. 2011. “Strengthening Chile’s Rule- Based Fiscal Framework.” IMF Working Paper 11 / 17. Washington, DC: International Monetary Fund.

Debrun, Xavier. 2011. “Democratic Accountability, Defi cit Bias and Independent Fiscal Agencies.” IMF Working Paper. Washington, DC: International Monetary Fund.

Debrun, Xavier, David Hauner, and Manmohan S. Kumar. 2009. “Independent Fis-cal Agencies.” Journal of Economic Surveys 23 (1): 48–81.

Debrun, Xavier, and Manmohan S. Kumar. 2007. “Fiscal Rules, Fiscal Councils and All That: Commitment Devices, Signaling Tools or Smokescreens?” In Fiscal Policy: Current Issues and Challenges, edited by Banca d’Italia, 479–512. Rome: Bank of Italy.

Eichengreen, Barry, and Charles Wyplosz. 1998. “The Stability Pact: Minor Nui-sance, Major Diversion?” Economic Policy 26:67–113.

European Commission. 2006. “Public Finances in EMU.” European Economy, 3.———. 2009. “Ageing Report.” European Economy, 2.Galí, Jordi, and Roberto Perotti. 2003. “Fiscal Policy and Monetary Integration in

Europe.” Economic Policy 37:533–72.Guichard, Stéphanie, Mike Kennedy, Eckhard Wurzel, and Christophe Andre. 2007.

“What Promotes Fiscal Consolidation: OECD Country Experiences.” OECD Economics Department Working Papers 553. Paris: OECD.

Hallerberg, Mark, Rolf Rainer Strauch, and Juergen von Hagen. 2009. Fiscal Gov-ernance in Europe. Cambridge: Cambridge University Press.

International Monetary Fund (IMF). 2005. “Promoting Fiscal Discipline—Is There a Role for Fiscal Agencies?” SM / 05 / 263.

———. 2009. Fiscal Rules—Anchoring Expectations for Sustainable Public Finances. Fiscal Affairs Department.

———. 2010. “A Historical Public Debt Database.” http: // www.imf.org / external / pubs / ft / wp / 2010 / data / wp10245.zip.


Jonung, Lars, and Martin Larch. 2006. “Improving Fiscal Policy in the EU. The Case For Independent Forecasts.” Economic Policy 47:491–534.

Kopits, George. 2001. “Fiscal Rules: Useful Policy Framework or Unnecessary Ornament?” IMF Working Paper 01 / 145. Washington, DC: International Mon-etary Fund.

Kopits, George, and Steven Symanski. 2001. “Fiscal Rules: Useful Policy Framework or Unnecessary Ornament.” IMF Working Paper WP / 01 / 145, Fiscal Affairs Department. Washington, DC: International Monetary Fund.

Krogstrup, Signe, and Charles Wyplosz. 2010. “A Common Pool Theory of Supra-national Defi cit Ceilings.” European Economic Review 54 (2): 273–81.

Kydland, Finn E., and Edward C. Prescott. 1977. “Rules Rather Than Discretion: The Inconsistency of Optimal Plans.” Journal of Political Economy 85 (3): 473–91.

Larch, Martin. 2010. “Fiscal Performance and Income Inequality: Are Unequal Societies More Defi cit- Prone? Some Cross- Country Evidence.” Unpublished Paper. European Commission.

National Conference of State Legislatures (NCSL). 2010. “State Balanced Budget Provisions.” NCSL Fiscal Brief, October.

Organization for Economic Cooperation and Development (OECD). n.d. Economic Outlook. http: // www.oecd.org / eco / economicoutlook.htm.

Persson, Torsten, and Guido Tabellini. 2000. Political Economics: Explaining Eco-nomic Policy. Cambridge, MA: MIT Press.

Pisauro, Giuseppe. 2003. “Fiscal Decentralization and the Budget Process: A Simple Model of Common Pool and Bailouts.” Italian Society for Public Economics Working Paper 294.

Poterba, James, and Kim Rueben. 1999. “State Fiscal Institutions and the US Munic-ipal Bond Market.” In Fiscal Institutions and Fiscal Performance, edited by J. Poterba and J. von Hagen, 181–208. Chicago: University of Chicago Press.

Roubini, Nouriel, and Jeffrey Sachs. 1989. “Political and Economic Determinants of Budget Defi cits in Industrial Democracies.” European Economic Review 33:903–38.

Schuknecht, Ludger, Juergen von Hagen, and Guido Wolswijk. 2008. “Government Risk Premiums in the Bond Market, EMU and Canada.” Working Paper 879, European Central Bank (ECB).

Tejedo, Maria. 2007. “State Fiscal Institutions: An Evolution.” PhD diss. University of Maryland at College Park.

Tellier, Genevieve, and Louis M. Imbeau. 2004. “Budget Defi cits and Surpluses in the Canadian Provinces: A Pooled Analysis.” Unpublished Paper. University of Ottawa.

von Hagen, Juergen. 2002. “Fiscal Rules, Fiscal Institutions, and Fiscal Perfor-mance.” The Economic and Social Review 33 (3): 263–84.

von Hagen, Juergen, and Ian J. Harden. 1994. “National Budget Processes and Fis-cal Performance.” European Economy Reports and Studies 3:311–408.

Wren- Lewis, Simon. 2011a. “Comparing the Delegation of Monetary and Fiscal Policy.” Discussion Paper 540. Department of Economics, Oxford University.

———. 2011b. “Fiscal Councils: the Office for Budget Responsibility.” CESifo DICE Report 3:50–53.

Wyplosz, Charles. 2005. “Fiscal Policy: Institutions versus Rules.” National Institute Economic Review 191:70–84.

———. 2006. “European Monetary Union: The Dark Sides of a Major Success.” Economic Policy 46:207–62.

This PDF is a selec on from a published volume from the Na ... · This PDF is a selec on from a...

Documents

Transcript of This PDF is a selec on from a published volume from the Na ... · This PDF is a selec on from a...